University of Illinois at Urbana-Champaign
Questioning the Information Subtest of the Wais-R: Dimensionality, DIF and DBF Analysis, and the Assessment of Category-Specific Semantic Memory Disturbances in Unilaterally Brain -Lesioned Patients
Abstract
dc:description"In this investigation the WAIS-R Information subtest was conceptualized as a measure of auditory-verbal semantic memory. The study's major purposes were to: (1) test the prediction that a Left Hemisphere Lesion group (LHL; n = 207) would perform significantly more poorly on the Information subtest than would a Right Hemisphere Lesion group (RHL; n = 203) and a Normal Control group (NC; n = 409), (2) test the prediction that the subtest's underlying dimensional structure would be multidimensional for all three examinee groups, and (3) investigate the predictions that either items or item clusters would manifest either differential item functioning (DIF) and/or differential bundle functioning (DBF) in interaction with the LHL group. With respect to this latter hypothesis, it was predicted that single items representing content known to be sensitive to left hemisphere lesion would exhibit DIF, and that DBF would emerge to demonstrate the LHL group would experience differential difficulty with item clusters reflecting the category-specific semantic domain of ""non-living/inanimate"" things and knowledge categories suggested by Kaplan and her colleagues (1991). Results showed the LHL group performed more poorly overall than the RHL and NC groups. In addition, the RHL group performed more poorly on this subtest than did the NC group, although the difference was small. Using as a theoretical basis the multidimensional item response theory (MIRT) model of Stout (1987, 1990) and its associated statistic, DIMTEST, the subtest was found to be multidimensional for all three examinee groups. In contrast, a factor analytic investigation of the subtest revealed a unidimensional underlying structure for all three groups. DIF and DBF were conceptualized using Shealy and Stout's (1993a, 1993b) MIRT model and analyses were convened using an associated statistic, SIBTEST. Various items and item clusters exhibited DIF/DBF in interaction with the LHL group. However, the ""non-living/inanimate"" things item subset did not exhibit DIF/DBF in interaction with this group, but instead exhibited DIF/DBF in interaction with the RHL group. Finally, various items and item clusters representing the semantic categories inspired by the proposals of Kaplan, et al. (1991) exhibited DIF and/or DBF. Results are discussed and the implications for practice and future research are offered."
Degree
thesis:*- Name thesis:degree_name
- Ph.D.
- Level thesis:degree_level
- Dissertation
- Discipline thesis:degree_discipline
- Education
- Grantor
- University of Illinois at Urbana-Champaign
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Chapman, Saul Howard
- Contributors dc:contributor
-
- Farmer, Helen S.
Subjects
dc:subject × 1Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- (MiAaPQ)AAI9944815
- OAI identifier oai:identifier
- oai:www.ideals.illinois.edu:2142/80315